378 DISCOVERY REPORTS 



high percentage of these females which are found to be pregnant (p. 455); this requires appreciably 

 more than one ovulation in 2 years. 



The second case (t = 2) is a possibility, but for a 2-year breeding cycle it gives a ratio of ovulation cor- 

 pora to pregnancy corpora of approximately 1 : 2. This does not fit the conclusions about the percentage 

 regression factor and the data on the size frequency of fully regressed corpora albicantia (p. 374) 

 which indicate a ratio of 2 : 1 ; nor is it in agreement with the evidence from the sexual cycle (p. 460 ). 



The first and simplest case (t = 1) appears much the most probable. For a 2-year breeding cycle 

 it gives a ratio of ovulations to pregnancies of approximately 2:1, which is in agreement with the 

 conclusions drawn from the size frequency of fully regressed corpora (p. 374). It is also in agreement 

 with estimates of the average annual rate of ovulation obtained by other quite independent methods 

 (p. 460) It also means that on average the regression to 'old' corpora albicantia takes about 3 years, 

 which was the opinion of Van Lennep (1950, p. 596), based on histological grounds. This evidence 

 strongly suggests that the average annual increment of corpora is about 1-5, although on the present 

 evidence a possible increment of 075 cannot be completely excluded. Further support for t = 1, 

 comes from the growth estimates (on pp. 413-415, and Text-figs. 38 and 39). 



In Text-figs. 21 and 22 the mean numbers of 'old' and recent (that is 'young' and 'medium') 

 corpora albicantia for different total corpora numbers have been plotted and curves fitted to them. It is 

 then apparent that the average number of recent corpora continues to increase with age. One possible 

 explanation of this is that the average annual production of corpora albicantia increases with age and 

 as this has a very important bearing on the question of age-determination it must now be dealt with. 



Other considerations being equal the formation of ' old ' corpora albicantia should lag about 3 years 

 (strictly three sampling periods) behind the production of ' young ' corpora, but should take place at 

 the same rate. The regression line describing the rate of accumulation of ' old ' corpora should there- 

 fore have a slope of 1 -o. In fact the regression line (calculated by the method of least squares and ignoring 

 the first six points) is described by y = — 3-433 + 0-885*, when y is the number of 'old' corpora 

 albicantia, and x is the total number of corpora albicantia. This gives an intercept on the x-axis at 

 3-88, suggesting that if 'young' corpora take three sampling periods to regress to 'old', then the 

 annual increment in the first 3 years after puberty averages about 1-3. We know that there are in the 

 material 88 females in their first pregnancy, and these had a mean number of 1-420 ±0-146 corpora, 

 which is in close agreement. The intersection of the two curves in Text-fig. 21 is at 8-9 corpora, when 

 4-45 recent corpora and 4-45 ' old ' corpora have formed. This implies that during the time it takes for 

 4-45 ' old ' corpora to accumulate a further 4-45 ' young ' and ' medium ' corpora are formed and that 

 ' old ' corpora are being added at about the same rate as ' young ' corpora are formed. This suggests 

 that ' young ' corpora form at the rate of about 1 -48 per year. 



The average number of recent corpora then increases from 4-45 at this intersection to 5-8 when 

 a total of 20 corpora have accumulated and to 7 when a total of 30 corpora have accumulated, 

 and so on. This apparently corresponds to increases in the annual rate of production of corpora 

 from 1-3 just after puberty to approximately 1-5, 1-9 and 2-3 at later ages. In terms of the number of 

 ovulations per 2-year cycle it implies a 50% increase over this age range from about 3 to over 4-5, that 

 is more than one extra ovulation per cycle. 



There is some evidence that the frequency of dizygotic twinning increases with age as a result of 

 multiple ovulations (Kimura, 1957), but only by a few per cent. Accessory corpora lutea amount to 

 only about 4% of all corpora lutea and multiple ovulations increase only slightly with age (see p. 454). 

 Nor is there any evidence for any other kind of increase in the ovulation rate. The percentage of 

 mature females with corpora lutea in the ovaries remains remarkably constant with increasing age 

 (Text-fig. 52). 



